Purpose: To determine whether intraarterial digital subtraction angiography (DSA) is as accurate as cut-film (film hard-copy) angiography (CFA) in the evaluation of suspected pulmonary embolus.
Materials and methods: Under a protocol approved by the institutional review board, CFA and DSA images were obtained in identical posteroanterior and oblique projections in one lung of each patient undergoing pulmonary angiography (n = 80). Diagnoses based on results of blinded review of each study (CFA vs DSA) by three separate reviewers (80 patients x three reviewers = 240 diagnoses for each modality) were compared with the diagnoses made by the physician who performed the procedure on the basis of CFA, DSA, and clinical data and with the consensus diagnoses obtained by means of group review of both studies together if necessary. The reviewers' confidence in their diagnoses was graded from 1 (certain) to 10 (uncertain).
Results: Pulmonary emboli were present in 13 (16%) of 80 patients. Two hundred thirty-six (98.3%) of 240 DSA diagnoses and 231 (96.3%) of 240 CFA diagnoses were correct. The sensitivity (correct identification of emboli by all three reviewers) of DSA was 92% and of CFA was 69% (P = .083). The specificities of the modalities were not statistically significantly different. The reviewers' confidence was significantly better in their DSA diagnoses than in their CFA diagnoses (P < .004).
Conclusion: DSA allows more confident detection of pulmonary embolus than does CFA, with no loss in diagnostic accuracy.